Assessment of reproductive status in animals generally depends on monitoring hormone concentrations in plasma, but blood sampling often involves significant stress to the subject. Monitoring steroid profiles by assaying excreted steroids in urine and/or fecal samples is non-invasive, but does pose some problems. Unlike plasma assays, urinary and fecal steroid analyses are of relatively little value in monitoring rapid, short-term changes in hormone concentrations (Heistermann et al., 1993) because there is a significant delay between production and excretion of steroids. However, such assays do enable measurement of "pooled" hormone concentrations over time (Wildt et al,. 1995;Brown, 1997).Hormones do not usually appear unchanged in the urine or feces, but are present as a range of metabolites produced by the liver, the major site of steroid metabolism (Kime, 1987). The proportions and forms of steroid excreted differ between the two excretory pathways, and may be highly variable between species (Aldercreutz et al,. 1980). In the maned wolf (Chrysocyon brachyurus) 97% of testosterone (T) is excreted as unconjugated metabolites in the feces (Vellosso et al., 1998), while in African wild dogs (Lycaon pictus) 60 % of steroids are excreted in the feces and 40 % in the urine (Monfort et al., 1997). A proportion of the excreted steroid is present as water-soluble conjugates -glucuronidates and/or sulfates -which can complicate hormone analysis. In most mammals, however, voided feces contain a higher percentage of free than conjugated steroids Vellosso et al., 1998;Adams et al., 1994), so excreted steroids are measurable by standard assay procedures. However, even if the same hormones are produced, different liver enzymes may produce different metabolites for excretion (Hodges, 1986); fecal steroid assays therefore need to be validated separately for each species.
1Most studies of fecal steroids have been conducted in mammalian species (Schwarzenberger et al., 1996), with limited work on birds such as the kakapo (Strigops habroptilus) (Cockrem & Rounce, 1995). This critically endangered, cryptic species is not amenable to invasive monitoring strategies, but Cockrem and Rounce (1995) have successfully utilized the ratio of fecal T to estrogen to assign sex in kakapo. Fecal steroid monitoring has also been used to follow the reproductive cycle in Japanese crested ibis (Nipponia nippon), and rock ptarmigian (Lagopus mutus) (Kikuchi & Ishi, 1997), and in the northern spotted owl (Strix occidentalis caurina) (Wasser et al.,1997;Brown et al., 1995).There are, however, only two published reports of the application of fecal steroid monitoring to reptiles. The reproductive cycles, characterized by behavioral observations and ultrasound imaging, of four species of tortoise (Geochelone elephantopus, G. gigantea;Testudo gracea, T. hermanni) have been positively correlated with profiles of fecal steroid metabolites (Casares, 1995;Dobeli et al., 1992). Ideally, fecal steroid profiles should be validated against plasma steroid profil...